In the case of the expression of heterologous proteins in prokaryotes, in the host cells these proteins often form inactive, sparingly soluble aggregates (so-called "refractile bodies") which, in addition, are also contaminated with proteins of the host cells. It is assumed that the formation of such "refractile bodies" is a result of the high protein concentrations in the cell arising in the case of the expression. It is known that, in the case of the formation of large amounts of enzymes in the cell, the aggregation of the enzymes to insoluble, high molecular, mostly inactive particles takes place. Before such proteins can be used, e.g. for therapeutic purposes, they must consequently be purified and converted into their active form.
According to known processes, a reactivation of such proteins present as aggregates takes place in several steps (cf. e.g. R. Jaenicke, FEBS Federation of European Biochemical Socities, Vol. 52 (1979) 187 to 198; R. Rudolph et al., Biochemistry 18 (1979) 5572 to 5575).
In the first step, a solubilisation is achieved by the addition of strong denaturing agents, for example guanidine hydrochloride or urea, in high concentration or by the addition of strongly acidic agents, for example glycine/phosphoric acid mixtures. As further adjuvants, there have proved useful reducing SH reagents (e.g. dithioerythritol, DTE) and EDTA, for example in the renaturing of LDH. Insofar as the protein is contaminated by proteins of the host cells, as next step there follows a purification with per se known and usual methods, e.g. gel or ion exchanger chromatography. Subsequently, it is highly diluted in order that the concentration of the denaturing agent becomes smaller. In the case of the use of guanidine hydrochloride, it is thereby diluted to values below 0.5 mole/l. In the case of enzymes with free SH groups, the addition of agents protecting the SH groups proves to be advantageous (cf. e.g. R. Jaenicke, Journal Polymer Science, Part C, 16 (1967) 2143 to 2160).
In EP-A-0114506, processes are described for the isolation, purification and reactivation of some heterologous expression products from bacterial cultures; for the reactivation, the solutions of the "refractile bodies" in a strong denaturing agent are
a) transferred directly into a solution of a weaker denaturing agent which is then subjected to oxidising conditions for the reformation of disulphide bridges;
b) the protein is sulphonated, then transferred into a solution in a weak denaturing agent and the S-sulphonate groups are converted into --S--S-- group by treatment with a sulphhydryl reagent in its reduced and oxidised from, e.g. with GSH/GSSG; or
c) the solution in a weak denaturing agent is treated directly with the sulphhydryl reagent, e.g. with GSH/GSSG. A typical example in which the above-discussed problems arises is t-PA.
The main component of the protein matrix of coagulated blood is polymeric fibrin. This protein matrix is dissolved by plasmin which is formed from plasminogen via activation by the so-called plasminogen activators, e.g. by t-PA (tissue-type plasminogen activator). The enzymatic activity of natural t-PA or of t-PA obtained gene-technologically from eukaryotes (catalytic activation of plasminogen to plasmin) is very low in the absence of fibrin or fibrin cleavage products (FCP) but can be substantially increased in the presence of these stimulators (by more than a factor of 10). This so-called stimulatability of the activity is a decisive advantage of t-PA in comparison with other known plasminogen activators, such as urokinase or streptokinase (cf. e.g. M. Hoylaerts et al., J. Biol. Chem., 257 (1982) 2912 to 2019; Nieuwenhiuzen et al., Biochemica et Biophysica Acta 755 (1983) 531 to 533). Therefore, the factor of the stimulatability with BrCN cleavage products is variously given in the literature and given a value of up to 35.
A t-PA-like, non-glycosilated product is also formed in genetically manipulated prokaryotes (after introduction of the c-DNA); however, such a product does not have the stimulatability of the activity of a t-PA from eukaryotes. It is conceivable that the reason for this is that the redox conditions in the prokaryote cell differ in such a way from eukaryote cells from which the gene originates that, ab initio, a non-active product is formed which, for example, could be due to the fact that numerous SS bridges which the natural active molecule contains are linked in a false way or are not even formed. However, for the therapeutic use of t-PA, there is necessary not only the enzymatic activity as such but, in addition, also its stimulatability. Regarding the fact that the prokaryote cells presumably do not provide the correct conditions in order to form the activity of eukaryotic proteins in the correct way, reference is made to other substances in the EMBO Journal 4, No. 3 (1985) 775 to 780.
According to EP-A-0093639, for the reactivation of t-PA, the cell pellets obtained from E. coli are suspended in 6 mole/l. guanidine hydrochloride, treated with ultrasonics, incubated and subsequently dialysed for four hours against a solution of tris-HCl (pH=8.0), sodium chloride, EDTA and Tween 80. After dialysis, it is centrifuged, whereby the plasminogen activator activity is to be found in the supernatant. t-PA renatured in this way is admittedly proteolytically active but shows no measureable stimulatability by BrCN cleavage products (BrCN-FCP) of fibrin according to the process described in J. H. Vereijen, Thromb. Haemostas., 48 (3), 260-269 (1982).
For the reactivation of denatured proteins, no generally useable process is known from the state of the art; this applies quite especially for t-PA because the native protein possesses a very complex structure; it contains a free thiol group and 17 SS bridges which theoretically can be linked in 2.2.times.10.sup.20 different ways, whereby only one structure corresponds to the native state. The processes known from the state of the art for the reactivation of t-PA admittedly lead to a proteolytically active t-PA but which shows no measurable stimulatability; an activation process which leads to stimulatable t-PA is not known.